The idea is simple: at the end of each game, the winning software strengthens its actions, while the loser weakens them—but both have also learned to evaluate their moves more efficiently.

In practice, one can introduce some degree of randomness in various ways: setting or updating the parameters at random, diversifying the order of the examples, adding some noise to the data, or using only a random fraction of the connections—all these ideas improve the robustness of learning.

LEARNING IS EXPLORING THE SPACE OF POSSIBILITIES

Alerting, which indicates when to attend, and adapts our level of vigilance. Orienting, which signals what to attend to, and amplifies any object of interest. Executive attention, which decides how to process the attended information, selects the processes that are relevant to a given task, and controls their execution.

By learning to read, for example, we have acquired an abstract concept of each letter of the alphabet, which allows us to recognize it in all its disguises, as well as generate new versions:

Another scenario that can lead to children losing interest is when curiosity is punished. A child’s appetite for discovery can be ruined by an overly rigid pedagogical strategy.

Episodic memory:

Attention is the gateway to learning: virtually no information will be memorized if it has not previously been amplified by attention and awareness.

Let students sleep. Sleep is an essential ingredient of our learning algorithm. Our brain benefits each time we sleep, even when we nap. So, let us make sure that our children sleep long and deep.